View Resource

Return to search results   |   Browse resources
Allomeric Variation. 2. Developmental Instability of Extreme Phenotypes

abstract1 (full description below): The American Naturalist, Vol. 120, No. 6. (Dec., 1982), pp. 765-786.
Allomeric Variation. PDF/Acrobat file     Michael E. Soule; Janine Cuzin-Roudy

In Soule (1982), a theory is proposed that might explain some patterns in the variation of morphometric (linear or volumetric) traits, in terms of the random events that occur during growth and development. Simply stated, this theory proposes that the randomness inherent in all developmental processes is damped in proportion to the size or number of components of the character. Therefore, the more complex a trait, in the former sense, the less 'variable it should be, other things being equal. A corollary of this theory is that the additive genetic contribution to the phenotype should increase as complexity increases, or, in other words, in a set of similar traits arranged in rank order of decreasing variability, the lower the rank (or variability), the greater the tendency of the phenotypes of individuals to accurately express their genotypes. This is because the strength of the genetic signal, relative to the random noise, is a positive function of complexity. One implication of this hypothesis is that as the coefficient of variation (CV)decreases in a set of homologous traits, extreme phenotypes are more likely to be produced by relatively homozygous genotypes rather than by random developmental phenomena. The reason is that complexity itself diminishes the developmental significance of stochastic events and, conversely, enhances the faithful representation of the genotype. Similarly, relatively heterozygous genotypes will be more likely to be near the mean of phenotypic distributions as complexity increases. If this trend is real, we would predict that individuals which are phenotypically extreme in the distribution of a morphometric trait with a low CV will have less developmental stability for that trait than will individuals nearer the mean, assuming that relatively homozygous individuals are inferior to relatively heterozygous individuals for developmental stability. The latter assumption is supported by several studies which show that developmental stability is inversely related to homozygosity (see citations). This paper presents the results of several tests of this prediction.